Evolution of a Cambrian active continental margin: the Delamerian-Lachlan connection in southeastern Australia from a zircon perspective

University of Wollongong Research Online Faculty of Science, Medicine and Health - Papers Faculty of Science, Medicine and Health 2013 Evolution of a Cambrian active continental margin: the Delamerian-Lachlan connection in southeastern Australia from a zircon perspective C L. Fergusson University of Wollongong, cferguss@uow.edu.au A P. Nutman University of Wollongong, anutman@uow.edu.au T Kamiichi University of Hiroshima H Hidaka University of Hiroshima Publication Details Fergusson, C. L., Nutman, A. P., Kamiichi, T. & Hidaka, H. (2013). Evolution of a Cambrian active continental margin: the Delamerian-Lachlan connection in southeastern Australia from a zircon perspective. Gondwana Research, 24 (3-4), 1051-1066. Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: research-pubs@uow.edu.au

Evolution of a Cambrian active continental margin: the Delamerian- Lachlan connection in southeastern Australia from a zircon perspective Abstract The Early Palaeozoic Ross Delamerian orogenic belt is considered to have formed as an active margin facing the palaeo-pacific Ocean with some island arc collisions, as in Tasmania (Australia) and Northern Victoria Land (Antarctica), followed by terminal deformation and cessation of active convergence. On the Cambrian eastern margin of Australia adjacent to the Delamerian Fold Belt, island arc and backarc basin crust was formed and is now preserved in the Lachlan Fold Belt and is consistent with a spatial link between the Delamerian and Lachlan orogens. The Delamerian Lachlan connection is tested with new zircon data. Metamorphic zircons from a basic eclogite sample from the Franklin Metamorphic Complex in the Tyennan region of central Tasmania have rare earth element signatures showing that eclogite metamorphism occurred at ~ 510 Ma, consistent with island arc passive margin collision during the Delamerian( Tyennan) Orogeny. U Pb ages of detrital zircons have been determined from two samples of Ordovician sandstones in the Lachlan Fold Belt at Melville Point (south coast of New South Wales) and the Howqua River (western Tabberabbera Zone of eastern Victoria). These rocks were chosen because they are the first major clastic influx at the base of the Ordovician Bengal-fan scale turbidite pile. The samples show the same prominent peaks as previously found elsewhere (600 500 Ma Pacific-Gondwana and the 1300 1000 Ma Grenville Gondwana signatures) reflecting supercontinent formation. We highlight the presence of ~ 500 Ma non-rounded, simple zircons indicating clastic input most likely from igneous rocks formed during the Delamerian and Ross Orogenies. We consider that the most probable source of the Ordovician turbidites was in East Antarctica adjacent to the Ross Orogen rather than reflecting long distance transport from the Transgondwanan Supermountain (i.e. East African Orogen). Together with other provenance indicators such as detrital mica ages, this is a confirmation of the Delamerian Lachlan connection. Keywords active, continental, margin, delamerian, lachlan, cambrian, connection, evolution, southeastern, australia, zircon, perspective Disciplines Medicine and Health Sciences Social and Behavioral Sciences Publication Details Fergusson, C. L., Nutman, A. P., Kamiichi, T. & Hidaka, H. (2013). Evolution of a Cambrian active continental margin: the Delamerian-Lachlan connection in southeastern Australia from a zircon perspective. Gondwana Research, 24 (3-4), 1051-1066. This journal article is available at Research Online: http://ro.uow.edu.au/smhpapers/1076

1 Evolution of a Cambrian active continental margin: the Delamerian Lachlan connection in southeastern Australia from a zircon perspective C.L. Fergusson a,*, A.P. Nutman a, T. Kamiichi 2, H. Hidaka 2 a School of Earth & Environmental Sciences, University of Wollongong, Wollongong, New South Wales 2522, Australia b Department of Earth and Planetary Systems Sciences, University of Hiroshima, 1-3-1 Kagamiyama, Higashi-Hiroshima 739-8526, Japan *Corresponding author, Telephone: +61 2 4221 3860 Fax: + 61 2 4221 4250 E-mail address: cferguss@uow.edu.au (C.L. Fergusson)

2 ABSTRACT The Early Palaeozoic Ross Delamerian orogenic belt is considered to have formed as an active margin facing the paleo-pacific Ocean with some island arc collisions, as in Tasmania (Australia) and Northern Victoria Land (Antarctica), followed by terminal deformation and cessation of active convergence. On the Cambrian eastern margin of Australia adjacent to the Delamerian Fold Belt, island arc and backarc basin crust was formed and is now preserved in the Lachlan Fold Belt and is consistent with a spatial link between the Delamerian and Lachlan orogens. The Delamerian Lachlan connection is tested with new zircon data. Metamorphic zircons from a basic eclogite sample from the Franklin Metamorphic Complex in the Tyennan region of central Tasmania have rare earth element signatures showing that eclogite metamorphism occurred at ~510 Ma, consistent with island arc passive margin collision during the Delamerian( Tyennan) Orogeny. U Pb ages of detrital zircons have been determined from two samples of Ordovician sandstones in the Lachlan Fold Belt at Melville Point (south coast of New South Wales) and the Howqua River (western Tabberabbera Zone of eastern Victoria). These rocks were chosen because they are the first major clastic influx at the base of the Ordovician Bengal-fan scale turbidite pile. The samples show the same prominent peaks as previously found elsewhere (600 500 Ma Pacific Gondwana and the 1300 1000 Ma Grenville Gondwana signatures) reflecting supercontinent formation. We highlight the presence of ~500 Ma non-rounded, simple zircons indicating clastic input most likely from igneous rocks formed during the Delamerian and Ross Orogenies. We consider that the most probable source of the Ordovician turbidites was in East Antarctica adjacent to the Ross Orogen rather than reflecting long distance transport from the Transgondwanan Supermountain (i.e. East African Orogen). Together with other provenance indicators such as detrital mica ages, this is a confirmation of the Delamerian Lachlan connection. Keywords: Arc continent collision Convergent margin Eclogite Lachlan Fold Belt Zircon 1. Introduction An issue with wider implications for the development of the East Gondwana Pacific margin is the relationship between the Lachlan and Delamerian fold belts in southeastern Australia (Cawood, 2005; Foster et al., 2005; Squire and Wilson, 2005; Cayley, 2011). The long-held opinion has been that these two orogenic belts developed adjacent to each other with the Delamerian Fold Belt formed by Neoproterozoic continental rifting followed by orogenesis involving collision and

convergent margin tectonism in the Delamerian( Tyennan) Orogeny, whereas the Lachlan Fold Belt developed with Cambrian oceanic island arc and backarc basement followed by widespread Ordovician turbidite deposition and island arc development (Cas, 1983; Powell, 1984; Glen, 2005). The source of the Ordovician turbidites is clearly continental, as indicated by the quartz-rich nature of sandstones in the succession. Some authors infer derivation of the sandstones from the Delamerian mountain chain (Turner et al., 1996; Fergusson and Tye, 1999), whilst others argue for a more distant source in the East African Orogen at the junction of West and East Gondwana (Fig. 1) (the Gondwana Supermountain of Squire et al., 2006; Williams and Pulford, 2008). The East African Orogen has been considered a significant source for early Palaeozoic sedimentary units in Gondwana including North Africa and adjacent parts of the Middle East (Figure 1) (Squire et al., 2006; Meinhold et al., 2013) and may also have been a source for late Cambrian to earliest Ordovician sedimentary rocks along the Tethyan margin of north India (Cawood et al., 2007). It has also been argued that many of the early Palaeozoic units of the Lachlan Fold Belt are allochthonous and have been transported by hundreds of kilometres or more along strike-slip faults (Willman et al., 2002; Glen, 2005; Glen et al., 2009). These interpretations can be tested by analysis and compilation of provenance data including paleocurrents, sedimentary petrology, and isotopic data including zircon and mica ages of detrital components (Cas, 1983; Ireland et al., 1998; Fergusson and Tye, 1999; Fergusson and Fanning, 2002; Squire et al., 2006; Williams and Pulford, 2008). Detrital zircon ages have been widely used to indicate provenance and for the Ordovician quartzose turbidites the detrital zircon age spectra are seemingly uniform with most ages in the ranges 600 500 Ma and 1300 1000 Ma. Similar spectra have also been widely found in early Palaeozoic sandstones throughout Gondwana (Veevers, 2000; Squire et al., 2006; Cawood et al., 2007; Williams and Pulford, 2008; Meinhold et al., 2013). In syntheses of global zircon ages these age ranges are associated with supercontinent formation. The 1300 1000 Ma range for granitoid ages and detrital zircons is associated with formation of Rodinia whereas the 600 500 Ma range reflects the final part of the Pan-African Orogeny during Gondwana assembly (Condie et al., 2009; Rino et al., 2008; Voice et al., 2011). Differences occur between detrital and igneous ages within continents; for example, no significant igneous peaks occur in Antarctica and Australia in the range 1500 1000 Ma whereas a large detrital peak is at 1060 Ma in Antarctica and at 1200 Ma in Australia (Condie et al., 2009, p. 237). How these are interpreted has important implications for provenance assessments. Thus does the absence of 1300 1000 Ma igneous ages in Antarctica reflect a sampling bias due to cover or is a distant source implied by the large detrital peak at 1060 Ma? Using detrital ages of zircons and other geochemical and isotopic data from sediment and sedimentary rocks it has been concluded that much of interior East Antarctica consists of an amalgam of Grenville and older cratons in a matrix of Pan-African fold belts (Veevers and Saeed, 2008, 2011). We examine the connection between these orogenic belts by acquiring chemical and isotopic data from zircons in three widely spaced samples. The first 3

4 sample is a ~510 Ma (Black et al., 1997) eclogite from the Franklin Metamorphic Complex of Tasmania and the other two samples are of sandstones from the base of the Ordovician turbidites of the Lachlan Fold Belt at the Howqua River of eastern Victoria and Melville Point on the south coast of New South Wales (Fig. 2). The study of the Tasmanian eclogite was undertaken to confirm if its zircons grew during collision-related eclogite facies metamorphism or whether they formed in some other event unrelated to high-pressure metamorphism. Although metamorphism at ~510 Ma has been shown by chemical U Th Pb monazite dating to be widespread in Tasmania (Berry et al., 2005) and therefore part of the Delamerian( Tyennan) Orogeny, it has been assumed rather than demonstrated that the ~510 Ma SHRIMP U Pb zircon age on eclogite in the Franklin Metamorphic Complex reflects the timing of high-pressure metamorphism. We have therefore undertaken rare earth element (REE) analysis of these zircons to see if they have a low HREE abundance and no negative Eu anomaly signature that is diagnostic of zircons grown during eclogite facies metamorphism (Rubatto, 2002). The study of the Ordovician turbidite samples was designed to provide a better constraint on the initiation and provenance of the extensive East Gondwana turbidite fan (Fergusson and Tye, 1999; Williams and Pulford, 2008). This was done by selecting the least mature Ordovician sandstones that occur at the base of the succession in the Howqua River in eastern Victoria and at Melville Point on the south coast of New South Wales. These samples are most likely to reflect the source in the earliest Ordovician, during the final phases of the Delamerian Orogeny (Foden et al., 2006). 2. Geological setting 2.1 Tasmanides of southeast Australia The Tasmanides of southeast Australia consist of the inner Delamerian Fold Belt occurring in southeastern South Australia, western Victoria and western Tasmania and the outer Lachlan Fold Belt in New South Wales, Victoria and northeastern Tasmania (Fig. 2). The Delamerian Fold Belt consists of thick rift basin successions of Neoproterozoic age overlying Paleoproterozoic to Mesoproterozoic basement of the Gawler and Curnamona Cratons in South Australia (Preiss, 2000). In western Victoria, northwestern New South Wales and Tasmania, mafic volcanic rocks and associated sedimentary successions have been interpreted as part of a volcanic passive margin that formed at 600 580 Ma (Direen and Crawford, 2003a, b). In Tasmania, the Delamerian Fold Belt consists of Neoproterozoic basement and basinal successions overthrust by early Cambrian ophiolite complexes containing boninites indicative of early island arc development and succeeded by a middle to late Cambrian post-collisional volcanic and sedimentary succession (the Mt Read Volcanics) (Berry and Crawford, 1988; Crawford and Berry, 1992). The ophiolite complex was emplaced in an island arc passive margin collision along an east-dipping subduction zone west of the island arc (Berry and Crawford, 1988). The basement Neoproterozoic quartzose siliciclastic successions contain abundant detrital zircons of 1800 1650 Ma

with less abundant 1450 1400 Ma ones and some samples containing 1300 1000 Ma ages indicating a likely early Neoproterozoic age for the succession (Berry et al., 2001; Black et al., 2004). The position of Tasmania in the Tasmanides prior to the Ordovician is problematic with some authors arguing for its initial location along the East Antarctic passive margin prior to the Ross Orogeny (Berry et al., 2008; Gibson et al., 2011). In western Victoria, the Delamerian Fold Belt consists of the Glenelg and Grampians Stavely zones that contain igneous and metamorphic rocks associated with the Delamerian Orogeny (Crawford et al., 2003). Boninitic and tholeiitic volcanic rocks have been documented in drill holes and the significance of widespread magnetic mafic rocks under cover has yet to be determined. They may represent thrust sheets, similar to those in Tasmania, that have been emplaced over the former late Neoproterozoic passive margin in an island arc passive margin collision (Crawford et al., 2003). In western New South Wales, the Delamerian Fold Belt is preserved in the Koonenberry belt but in contrast to further south shows the development of a short-lived east-facing Delamerian active margin with an arc, forearc basin and eastern subduction complex (Greenfield et al., 2010, 2011), without an island arc passive margin collision. In contrast to the Delamerian Fold Belt, the Lachlan Fold Belt has an early to late Palaeozoic history. The tectonic development of the Lachlan Fold Belt is a widely debated issue. Numerous suggestions have been given for the Ordovician tectonic development including formation of multiple subduction zones (Gray and Foster, 2004), strike-slip emplacement of continental margin terranes (Glen et al., 2009) and island arc passive margin collision somewhat akin to the Delamerian Orogeny in Tasmania (Aitchison and Buckman, 2012). In the Cambrian, an oceanic realm developed resembling the modern western Pacific Ocean with island arcs, backarc basins and boninitic-tholeiitic successions (Crawford et al., 2003). In contrast, the Melbourne Zone is underlain by a northward continuation of the basement rocks of western Tasmania called the Selwyn block (Cayley et al., 2002; Cayley, 2011; Moore et al., 2013) as shown by magnetic connections across Bass Strait. Western Tasmania is thought to have reached its present location with respect to East Gondwana by the Early Ordovician (Li et al., 1997). In the earliest Ordovician, widespread quartzose turbidite deposition spread across much of the Lachlan Fold Belt following its earlier deposition in the Stawell Zone in the middle to late Cambrian (VandenBerg et al., 2000). Additionally, an island arc (Macquarie arc) was initiated in the east (Glen, 2005). The turbidite succession formed mainly through the Early to Middle Ordovician and in the eastern part of the fold belt was followed by widespread black shale and chert deposition (VandenBerg and Stewart, 1992). The relationship between the Macquarie arc and the surrounding Ordovician quartz turbidites are problematic with several explanations offered including juxtaposition by strike-slip faulting (Glen et al., 2009), island arc rotation and development of adjoining subduction complexes (Fergusson, 2009), and overthrusting of a passive margin by an island arc from the east (Aitchison and Buckman, 2012). For the Early Ordovician, a connection between the Gondwana margin and the Macquarie Arc has been argued by Glen et al. (2011) who showed that the Mitchell 5

6 Breccia at the base of the Macquarie Arc succession contains a detrital zircon age pattern similar to that documented for the Ordovician quartz turbidites. The younger part of the Macquarie arc succession has a reduced number of detrital inherited zircons and it is unclear if this reflects a real trend or is due to an increasing proportion of magmatic zircons present in the samples (Glen et al., 2011, p. 682). In the Late Ordovician widespread deformation, the Benambran Orogeny, caused crustal thickening in much of the central and eastern parts of the Lachlan Fold Belt into continental thicknesses such as in the Wagga Omeo Zone with intrusion of abundant Silurian granites (Ickert and Williams, 2011). This new phase of magmatic activity has been related by Fergusson (2003, 2009) to migration of Macquarie arc magmatism westwards into the adjoining subduction complex. In the mid Silurian to Middle Devonian much of the Lachlan Fold Belt was in an extensional setting with widespread silicic igneous activity, thick successions of deep to shallow marine rocks and intermittent episodes of contractional deformation associated with an east-dipping subduction zone at the eastern margin of the belt (Powell, 1984; Collins, 2002). 2.2 Franklin Metamorphic Complex The Tyennan block of central and southern Tasmania consists mainly of low-grade phyllite and quartzite with several higher grade complexes including the Franklin Metamorphic Complex. The Franklin Metamorphic Complex contains schist, minor quartzite, and lenses of metabasic rocks some of which are of eclogite grade (Meffre et al., 2000). These basic eclogites reached maximum metamorphic conditions at ~600 650 C and 1.5 GPa (Palmeri et al., 2009) and contain zircons with U Pb ages of 511±8 Ma (Black et al., 1997; Turner et al., 1998). The schists have widely developed garnet and kyanite and therefore have most likely been metamorphosed at depth with the associated eclogites (Meffre et al., 2000). 2.3 Howqua River The Howqua River crosses the contact between the Melbourne and Tabberabbera zones in eastern Victoria. The Melbourne Zone has a simply folded, upright Silurian to Early Devonian turbidite succession with a structurally complex belt further east (Mount Useful Fault Zone) that has windows containing middle to late Cambrian volcanic rocks of intermediate to silicic calc-alkaline affinity below a multiply deformed Late Ordovician black shale and Silurian turbidite succession (VandenBerg et al., 1995, 2006). Further east, the Tabberabbera Zone is dominated by an Ordovician and early Silurian turbidite succession overlain by Late Ordovician black shale (VandenBerg et al., 2000). In the Howqua River region, a largely complete Ordovician succession is developed with its basal earliest Ordovician turbidites underlain by bedded cherts which overlie a thick Cambrian altered mafic volcanic succession with upper tholeiitic and lower boninitic units (Fergusson, 1998; Crawford et al., 2003). Between the Cambrian volcanic rocks and the Mount Useful Fault Zone lies a structurally complicated assemblage of mélange containing chert, sandstone,

7 mudstone and blueschist blocks with Late Ordovician ages based on poorly preserved graptolites in low-grade metasedimentary rocks, a U Pb age on titanite and an Ar Ar age on mica in blueschist (Spaggiari et al., 2002a). The Howqua River section is particularly significant as it is one of the few places in southeastern Australia where a basal contact to the widespread Ordovician turbidite succession is exposed, with reliably-dated Tremadocian early Arenig graptolites (Fergusson, 1998). In contrast to the bulk of the overlying Ordovician turbidites, these basal sandstones have lower quartz (43 66%), higher feldspar (7 18%) and lithic clast (8 15%) contents (Fergusson and Tye, 1999). The tectonic interpretation of the Howqua River region and the Lachlan Fold Belt is a matter of continuing controversy. In the mid 1990 s, the idea of a double divergent subduction zone in central Victoria was introduced by Gray (1997) and Gray and Foster (1997) with subduction occurring towards the northeast generating the Tabberabbera Zone by subduction accretion of the thick turbidite succession and forming the Wagga Omeo metamorphic belt in an arc setting. These ideas have been vigorously debated but more recent confirmation of the subduction-accretion related structural style of the Tabberabbera Zone has been provided by Watson and Gray (2001) and Willman et al. (2005). 2.4 Melville Point Melville Point occurs on the south coast of New South Wales south of the southern end of the Permo Triassic Sydney Basin. At Melville Point, the base of the Ordovician turbidites is exposed in a coastal cliff and rock platform (Fig. 2b). It is conformable on a succession of chert and mudstone that is underlain by altered basalt (Prendergast, 2007). The Ordovician turbidites are part of the Adaminaby Group whereas east of Melville Point the Wagonga Group is exposed. It is a package of Cambrian and Ordovician mafic volcanic rocks, limestone, quartz-rich turbidites, mudstone and mélange. Altered basalts, similar to those at Melville Point, are widespread in the Wagonga Group and have been shown to have an ocean island basalt magmatic affinity (Prendergast and Offler, 2012). Structures in the Wagonga Group are consistent with its interpretation as part of a subduction complex (Prendergast, 2007). The turbidites at Melville Point consist of thick, massive beds of quartz-feldspar-lithic sandstone interbedded with thin-bedded sandstone and mudstone containing thin chert layers. A layer of thin chert near the western end of the rock platform contains conodonts of late Cambrian to earliest Ordovician age (Glen et al., 2004, p. 873), thus showing that the underlying chert-mudstone and mafic volcanic rocks are of probable Cambrian age. Middle to late Cambrian limestones occur associated with mafic breccias located ~3.5 km east in the Wagonga Group at Burrewarra Point (Bischoff and Prendergast, 1987). 3. Zircon petrography, dating and rare earth element chemistry 3.1 Analytical methods

8 Zircon concentrates were prepared by standard heavy liquid and magnetic separation techniques at the mineral separation laboratory of the Research School of Earth Sciences, the Australian National University (ANU). The Howqua River and Melville Point sandstone zircon concentrates were hand picked under a binocular microscope, and the selected grains were cast in epoxy resin discs together with the zircon Temora 2 reference material (Black et al., 2003). The discs were ground to reveal mid-sections through the grains and then polished. The grains were documented with reflected and transmitted light photomicrographs and by cathodoluminescence (CL) imaging. CL images of representative zircons are shown in Fig. 3. The Franklin metamorphic zircons had been U Pb dated in the 1990s (Black et al., 1997) and were located in the Geoscience Australia Mount Z1740, together with the U Pb calibration zircon SL13. The mount was loaned by Geoscience Australia for additional CL-imagery guided U Pb dating and REE analysis (Fig. 4). Zircons were U Pb dated on the SHRIMP II instruments at the University of Hiroshima and Geoscience Australia (Canberra). Analytical protocols followed Williams (1998), and reduction of the raw data used the ANU software PRAWN and Llead. 206 Pb/ 238 U of the Howqua River and Melville Point unknowns were calibrated using measurements of the Temora 2 reference material (U Pb ages concordant at 417 Ma; Black et al., 2003). 206 Pb/ 238 U of the Franklin metamorphics zircons were calibrated using SL13 zircon. U and Th abundance was calibrated using measurement of the reference material SL13 (U = 238 ppm). The zircon U Th Pb data are summarised in Table 1. The reduced and calibrated data were assessed and plotted using the ISOPLOT program of Ludwig (2003). Rare earth element analyses of some Franklin metamorphics zircons were undertaken using the Hiroshima University SHRIMP II instrument, in the normal mass-resolution mode (~6000), and according to the method outlined by Maas et al. (1992) and Hidaka et al. (2002). Abundances were calibrated using count rates obtained from analyses of the zircon standard 91500 (Wiedenbeck et al., 2004). Wiedenbeck et al. (2004) noted elemental heterogeneities within 91500. Given this observation, even if the very low abundances of La and Nd in our unknowns have been underestimated by an order of magnitude, this will not change the strongly fractionated, light REE depleted, heavy REE enriched (chondrite normalised) signatures of these zircons. REE analyses are summarised in Table 2. 3.2 Franklin metamorphics zircon morphology and U-Pb dating Franklin metamorphics eclogite sample 9322 0004 yielded small (mostly <100 µm long) zircons of equant to stubby prismatic habit. In CL images their internal structure is complex, with domains of sector zoning, irregular patches and recrystallisation being more common than oscillatory zoning. A few grains have distinct narrow rims that appear homogeneous and bright in CL images (grain 8, Fig. 3). These features are typical of zircons grown during eclogite facies metamorphism (Nutman et al., 2008). Fourteen U Pb analyses were undertaken on 9 grains (Table 1).

9 All sites have low Th/U (0.031 to 0.003), typical of metamorphic zircons. They all have close to concordant and statistically-indistinguishable U Pb ages (Fig. 4). However, analysis 8.1 of a distinct rim overgrowth has a different REE pattern (see below) and thus was not included in the weighted mean 206 Pb/ 238 U age of 504±7 Ma (95% confidence, MSWD=0.41) for the other 13 analyses. These analyses included all varieties of sites within the grains (sector zoned, patchy and recrystallised). Anomalous site 8.1 on a narrow rim yielded a 206 Pb/ 238 U age of 460±35 Ma (1σ), whereas core analysis 8.1 gave an age of 510±14 Ma. Although these ages are within error of each other, the petrographic relationships indicate that the rim is younger. 3.3 Franklin metamorphics zircon REE chemistry Five REE analyses were undertaken on the dated Franklin metamorphics eclogite sample 9322 0004 zircons: one on the distinct rim 8.1, and the others on the interiors of the grains with variable texture (2.2, 3.1, 5.1 and 6.2; Table 2). The latter four analyses of the 504±7 Ma population show a classic chondrite normalised REE pattern for zircons grown during eclogite facies metamorphism, with depressed HREE abundances (50 10 times chondrite) and devoid of any Eu anomaly (Fig. 4). For a rock like the gabbroic/doleritic protolith of sample 93222 0004, this pattern indicates zircon growth under eclogite facies, because plagioclase was absent (hence no negative Eu-anomaly) and there was abundant garnet present (resulting in the depressed HREE; Rubatto, 2002; Gilotti et al., 2004; Nutman et al., 2008). The pattern for rim analysis 8.1 does not display a negative Eu anomaly, but there is a steady increase in the HREE relative to reference chondrite, rather than the flat pattern seen for the other analyses. This suggests its growth in a slightly different environment (some garnet breakdown?) at the end of the eclogite facies metamorphic event. 3.4 Howqua River and Melville Point zircon U Pb dating Both the sample HQ9 from Howqua River and sample MVP from Melville Point yielded morphologically-similar zircon populations with the same age spectra, thus they are considered together here. CL images of representative zircons from the sample MVP are shown in Fig. 5. The grains all show abraded exteriors, thus they are devoid of in situ growth of metamorphic zircon. This is in keeping with the low grade (maximum greenschist facies) overprint on these rocks (Offler et al., 1998; Spaggiari et al., 2002b). The majority of the grains are oval or somewhat rounded prisms, whereas a minority of the grains in both samples are almost euhedral prisms. The rounded grains tend to show complex internal structures in CL images. Some have distinct cores and overgrowths, and truncation of zonation at the margins is common. Thus for example MVP grain 3 displays a Mesoproterozoic core, a broad Neoproterozoic rim, plus an outer thin partial rim, too narrow to be dated with a SHRIMP analytical spot (Fig. 5). For other rounded grains, zoning in old components have been truncated by abrasion at the grain boundary, such as for Neoarchean MVP

10 grain 16 (Fig. 5). These grains give Precambrian ages, particularly latest Neoproterozoic (600 550 Ma), Mesoproterozoic (1100 1000 Ma), with a lesser number of older grains present (Table 1; Fig. 6). The rarer more euhedral, prismatic grains show oscillatory zoning parallel to the grain margins, but generally are dark in the CL images, due to higher U and Th content. These grains give the youngest ages (Table 1; Fig. 6), and will provide the best constraint on the maximum age of deposition of the Melville Point and Howqua River sandstones. Therefore SHRIMP analyses were focussed on these younger grains, with repeat analyses undertaken on several of them, in order to establish their age(s) with greater confidence. However, the commonly high U and Th content means that modern loss of radiogenic Pb in these grains can be a problem, giving apparent 206 Pb/ 238 U ages younger than the 207 Pb/ 206 Pb ages (with the latter being less precise at the early Phanerozoic). Thus high U (>500 ppm) grains and sites gave the youngest, scattered ages, which is regarded as due to later isotopic disturbance in grains with the most structural damage from radioactive decay. Grains with less U (300 100 ppm) yielded concordant ages at ~500 Ma. Assessments of the ages of these grains are as follows: for HQ9 505±34 (2σ), 499±17 (95% confidence), 496±32 (2σ), 495±19 (95% confidence), 495±34 (2σ) Ma and for MVP 510±10 (95% confidence), 484±15 (95% confidence) Ma. (2σ used when only 1 analysis was undertaken on a grain.) This suggests a maximum depositional age of lowermost Ordovician (Tremadocian) for the turbidite fan in the eastern Lachlan Fold Belt. The remaining signature of older grains in these sandstones is classic eastern Gondwanan, with a predominance of 600 550 and 1100 1000 Ma grains (Fig. 6). 4. Discussion The main issue to be considered is the development of the Delamerian Fold Belt and its relationship to the deposition of a vast turbidite fan in the Lachlan Fold Belt. Earlier ideas on the Cambrian Ordovician paleogeography of southeastern Australia regarded the Delamerian Orogeny as a major episode of shortening, which thickened crust of the Delamerian Fold Belt and thereby provided a repeatedly-recycled source for deposition of the widespread siliciclastic turbidite fan (Turner et al., 1996; Fergusson and Tye, 1999). Various plate tectonic scenarios have been given for the Delamerian Orogeny including an island arc passive margin collision as exemplified by the geology of western Tasmania (Berry and Crawford, 1988; Meffre et al., 2000). Some recent work has raised questions about the vast Ordovician turbidite fan having formed adjacent to the Delamerian Fold Belt. It has been suggested that the deep-marine Ordovician siliciclastic rocks of the Lachlan Fold Belt formed one or more allochthonous terranes that have been transported along the East Gondwana margin (Baillie, 1985; Glen, 2005; Glen et al., 2009). Others suggested that long-distance sedimentary transport of sediment from the East West Gondwana Pan-African collision zones thousands of kilometres away provided an explanation for the detrital U Pb zircon ages in the Ordovician and other siliciclastic successions (Squire et al., 2006; Maidment et al., 2007; Williams and Pulford, 2008).

11 4.1 Delamerian( Tyennan) Orogeny From the space-time diagram for selected regions in southeastern Australia (Fig. 7) it is clear that the Delamerian ( Tyennan) Orogeny has substantial variations in terms of the timing and nature of events. Numerous intermittent episodes of rifting are associated with the formation of the Adelaide Rift Complex in South Australia dating from ~800 Ma onwards (Preiss, 2000). A 600 580 Ma passive margin has been recognised along the East Gondwana margin in the Delamerian Fold Belt (Direen and Crawford, 2003a, b). Widespread magnetic anomalies associated with volcanic successions with tholeiitic to alkaline magmatic affinities in the Koonenberry belt, western Victoria and in Tasmania have been interpreted as equivalent to the thick mafic volcanic units of seaward-dipping reflector sequences in volcanic passive margins. A marked change in provenance has accompanied the Delamerian Orogeny as shown by the U Pb detrital zircon ages published by Ireland et al. (1998). Prior to ~545 Ma, detrital zircon ages show that the Adelaide Rift Complex had various sources, such as a late Mesoproterozoic zircon source reflected in samples from the Bonney Sandstone and Marino Arkose, whereas other samples, such as the Niggly Gap beds and Mitcham Quartzite, were derived mainly from the adjoining Gawler craton. In contrast, a sample from the Kanmantoo Group shows a substantial 700 500 Ma age peak reflecting the input from a new source. Samples from the Normanville Group, the shallow marine succession underlying the Kanmantoo Group, reflect mainly the pre-545 Ma sources with derivation from the late Mesoproterozoic and 2000 1500 Ma peaks represented in the Heathersdale Shale sample and the 1830 Ma peak in the Mt Terrible Formation sample. In Antarctica, development of a passive margin along the Ross Orogen has been placed at ~700 Ma with initiation of convergence along the margin occurring at ~580 Ma, as shown by the incoming detritus containing abundant detrital zircon and muscovite with ages of 580 500 Ma (Goodge et al., 2002, 2004a, b). The convergence is related to the inception of a continental margin magmatic arc reflecting conversion of the passive to an active margin with more complicated events in Northern Victoria Land related to accretion of an island arc represented by the Bowers Supergroup (Gibson and Wright, 1985). New radiometric ages from the Delamerian Fold Belt in South Australia indicate an earlier start to convergence at ~545 Ma rather than at 514 Ma as previously considered (Turner et al., 2009). These authors further point out that the Normanville and Kanmantoo groups did not form in an extensional basin setting as part of the former passive margin but were part of a synorogenic basinal setting like the Byrd Group in Antarctica. In northwestern New South Wales, a convergent margin assemblage is recognised for the Cambrian Mount Wright Volcanics and associated shallow-marine successions formed in arc to forearc environments with a deep-marine turbidite succession to the northeast (Teltawongee Group) considered the equivalent subduction complex (Greenfield et al., 2011). U Pb SHRIMP ages show that this assemblage lasted for 20 million years at ~525 505 Ma. It was succeeded by intense

compressional deformation in the Delamerian Orogeny at 505 500 Ma in contrast to its longer time interval in South Australia. The arc-forearc assemblage developed over the former passive margin succession. One sample from the Nundora Formation in the western Teltawongee Group shows ages of detrital zircons with a prominent peak at 1178±17 Ma and no ages in the interval 600 500 Ma, whereas another sample from the Cooper Mine Range Formation shows a significant peak at 700 500 Ma (Greenfield et al., 2010, p. 346) indicating the varied provenance of units prior to the Delamerian Orogeny. The post-delamerian succession in the Koonenberry belt shows a similar detrital zircon age signature as found in the Kanmantoo Group with prominent age spectra in the range 600 500 Ma and 1300 1000 Ma as shown by samples from the Mutawintji Group and the Warratta inlier. However, in detail both these samples show prominent peaks in zircon ages at ~580 Ma consistent with derivation from the late Neoproterozoic rifted volcanic succession including the Mt Arrowsmith Volcanics (Fig. 8a) (Greenfield et al., 2010, p. 346; Greenfield et al., 2011). Additionally, one sample is from the Bilpa Conglomerate, which is poorly sorted with clasts up to 1 m across that are dominantly metasandstone, metapelite and mafic volcanic rocks indicative of local basement rock types affected by the Delamerian Orogeny (Pahl and Sirorska, 2004; Greenfield et al., 2010). The Bipla Conglomerate is part of a coarse-grained delta deposit formed in continental and shallow marine environments (Pahl and Sirorska, 2004). Thus we contest interpretations that these Pacific-Gondwana detrital zircon signatures, as found in the matrix of the Bilpa Conglomerate, are indicative of long-distance sedimentary transport (see below; Squire et al., 2006; Williams and Pulford, 2008). In western Victoria, substantial cover by younger units has obscured the geological relationships, especially in the eastern part of the Delamerian Fold Belt. The Delamerian Orogeny in western Victoria is associated with development of abundant plutonic rocks in the Glenelg Zone (Crawford et al., 2003). Collision of an island arc boninitic-tholeiitic assemblage with a rifted margin has been postulated and was followed by post-collisional igneous activity, similar to events in Tasmania (VandenBerg et al., 2000; Crawford et al., 2003). It has been argued that the Glenelg Zone represents part of the passive margin succession that has been accreted into an accretionary wedge at an east-dipping subduction zone associated with the boninitic island arc, similar to relationships proposed for Northern Victoria Land (Gibson et al., 2011). These authors argued that following collision of the east-directed subduction zone with the passive margin, a new west-dipping subduction zone initiated east of the accreted island arc and caused the widespread igneous activity in the Glenelg Zone in addition to the post-collisional igneous activity of the Grampians Stavely Zone. Ar/Ar ages from the Moorambool Metamorphic Complex in the western Stawell Zone indicate exhumation of garnet amphibolites at ~500 Ma from this west-dipping Delamerian subduction zone (Miller et al., 2005). Cayley (2011) has reinterpreted the tectonic development of the Delamerian Fold Belt in western Victoria as an active east-facing margin developed in the Early to Middle Cambrian with intensification of deformation in the Late Cambrian associated with collision of an outboard 12

microcontinent, Van Dieland, that includes Tasmania and its inferred northern subsurface continuation, the Selwyn Block forming a deeply buried metamorphic basement to much of central Victoria (Cayley et al., 2002). In Tasmania, the Tyennan Orogeny is attributed to an island arc passive margin collision with emplacement of an island arc ophiolite allochthon of Cambrian boninitic-tholeiitic rocks over a Neoproterozoic passive margin (Berry and Crawford, 1988; Crawford and Berry, 1992). Metamorphic complexes, such as the Franklin Metamorphic Complex, were formed by subduction of a thin passive margin, lacking a young thick sedimentary cover, under the island arc as has occurred in Oman (Meffre et al., 2000). These authors also suggested that exhumation of these metamorphic rocks was aided by extensional unroofing associated with formation of the post-collisional Dundas Trough. The location of Tasmania prior to and during the Delamerian( Tyennan) Orogeny has also been widely debated in the literature. Paleomagnetic data from northwest Tasmania imply that in the Early Ordovician western Tasmania was in its approximate present position with respect to East Gondwana (Li et al., 1997). From monazite ages indicating metamorphism and granitoid intrusion in the interval 1290 920 Ma, Berry et al. (2008) inferred that the west Tasmania terrane, including the South Tasman Rise, was most likely connected to the Transantarctic Mountains prior to rifting at ~580 Ma. Arguing from relationships in western Victoria and Northern Victoria Land and including constraints from the Southern Ocean passive margins, Gibson et al. (2011) developed a comprehensive model for the Ross Delamerian Orogeny. They suggested that a common east-dipping subduction zone existed with eastwards subduction of the East Gondwana passive margin under an island arc from north to south in western Victoria, Northern Victoria Land and Tasmania. Collision resulted in flipping of the subduction zone with development of west-dipping subduction and magmatic activity during the Middle to Late Cambrian. Tasmania has therefore been displaced from an original position near South Victoria Land along major sinistral strike-slip faults, such as the Avoca Fault and it offshore southern continuation. Somewhat less strike-slip displacement of the west Tasmania terrane and the Selwyn Block was argued by Cayley (2011). Just how any largescale sinistral strike-slip transport is accommodated in mainland southeastern Australia is problematic. In summary, the Delamerian Orogeny of southeastern Australia reflects initiation of a continental margin convergent assemblage from 545 Ma but most widely developed in the interval 520 505 Ma with emplacement of island arc allochthons in the later part of this interval. The context of west Tasmania along the continental margin is poorly resolved but with a possibility that it may have collided with the Delamerian active margin at ~505 Ma prior to formation of a post-collisional volcanic succession developed in the Mt Read Volcanics of western Tasmania, the Mt Stavely Complex and equivalent units in western Victoria and the Licola Volcanics along the eastern margin of the Melbourne Zone in central Victoria (Cayley, 2011). Alternatively, others have argued for a more remote location for Tasmania further southwards along the Ross Orogen prior to terrane displacement that may have 13

14 involved considerable strike-slip displacement to its Ordovician setting in East Gondwana (Berry et al., 2008; Gibson et al., 2011). The Delamerian convergent assemblage is unusual in several ways. It is characterised by the development of low to locally high-grade metamorphic complexes associated with granitoids but lacks widespread granodioritic-tonalitic plutonic rocks more typical of an eroded continental margin arc. These are thought to occur along the western margin of the Ross Orogen but are presently covered by the East Antarctic ice sheet. Their presence has been inferred by abundant 600 500 Ma detrital zircons in siliciclastic rocks of the exposed part of the orogen (Goodge et al., 2002, 2004a, b; see below). For southeastern Australia, the connection between the Ross and Delamerian orogens is complicated by island arc collision in western Tasmania and suggested for western Victoria. We favour the idea of Cayley (2011) that the Delamerian Orogeny at 505 500 Ma was caused by collision of west Tasmania, and associated fragments, accompanied by development of new island arc systems in the paleo-pacific Ocean such as the Macquarie Arc (Fig. 9). 4.2 Potential sources of the Ordovician turbidites? Paleogeographic reconstructions assuming the present-day spatial arrangements between orogenic belts show that the source of the Ordovician turbidites was in the Delamerian Fold Belt to the west and south (Cas, 1983). This is consistent with available paleocurrent data indicating westward derivation for parts of the turbidite fan (Powell, 1984; Fergusson and VandenBerg, 2003). A source in the Delamerian Fold Belt is also consistent with sedimentary petrography of sandstones in the succession with abundant quartz, minor feldspar and minor metamorphic detritus (Fergusson and Tye, 1999). Sr isotopic data from whole rock samples of the Ordovician turbidites are also appropriate for sources such as the Kanmantoo Group in the Delamerian Fold Belt and igneous rocks in western Tasmania (Gray and Webb, 1995). Detrital mica ages within the Ordovician sandstones have been determined in several Ar/Ar isotopic studies. Sandstone samples from the southern Bendigo Zone and the central part of the Tabberabbera Zone have detrital muscovite ages in the range 512 480 Ma (Turner et al., 1996). From sandstones in the Stawell and Bendigo zones, Foster et al. (1998) showed muscovite ages range 580 470 Ma with most ages in the interval 510 500 Ma. These data are in agreement with the timing of peak metamorphism in the Delamerian Fold Belt (Foden et al., 2006; Turner et al., 2009) and provide further support for a provenance connection between the Delamerian mountain chain and the Ordovician turbidites. The source of the vast Ordovician turbidite fan of the Lachlan Fold Belt presents a dichotomy when the detrital zircon U Pb ages are taken into account because of the lack of an obvious source of the 600 500 Ma and 1300 1000 Ma zircons. The samples from the Howqua River and Melville Point are considered informative given that these sandstones are earliest Ordovician and the least mature compositions recorded from the Ordovician turbidites, with higher feldspar and lithic

15 contents (Fergusson and Tye, 1999). These samples show the common detrital zircon age spectra determined from other Ordovician sandstones with significant age intervals at 600 500 Ma and 1300 1000 Ma (Fig. 8b-g) (Ireland et al., 1998; Williams, 1998; Fergusson and Fanning, 2002). This signature has been widely recorded from sandstones in Australia. For example, two samples from the Triassic Hawkesbury Sandstone in the Sydney region are dominated by the 700 500 Ma detrital zircons, considered the Pacific Gondwana igneous component, and regarded as exotic and derived from East Antarctica (Sircombe, 1999). Alternatively, several authors have related these detrital zircon signatures to long-distance transport of significant volumes of sediment from the East African Orogen as discussed in the next section (Squire et al., 2006; Maidment et al., 2007; Williams and Pulford, 2008). 4.3 Is the East African Orogen the only possible source for the Ordovician turbidites? It has been argued that the widespread detrital zircon age ranges of 1200 900 Ma and 650 550 Ma in Cambrian to Ordovician quartz-rich sandstones in different parts of Gondwana reflect erosion of the Transgondwanan Supermountain which was formed from the East African Orogen during collision of West and East Gondwana (Squire et al., 2006). Widespread deposition of mature quartz-rich sandstones of Cambrian to Ordovician ages in North Africa, the Middle East and displaced terranes in Europe have been related by Meinhold et al. (2013) to erosion and rejuvenated uplift in parts of the Transgondwanan Supermountain. It is also possible that the similar detrital ages reported by Cawood et al. (2007) in early Palaeozoic sandstones along the Tethyan margin of India are another component of this Gondwana super-fan derived from the Transgondwanan Supermountain. In the Ross Orogen and Delamerian and Lachlan fold belts similar signatures are common in quartz-rich sandstones and include the Ordovician turbidites which Squire et al. (2006) argued were part of the Gondwana super-fan derived by very long distance sedimentary transport from the southern part of the East African Orogen. As noted above a similar signature also occurs in Triassic quartz-rich sandstones of the Hawkesbury Sandstone in the Sydney Basin which indicates a markedly long-term uniformity of the apparent long-distance source. We argue that such long-distance transport suggested for the Ordovician turbidites is unlikely. Does the Transgondwanan Supermountain (i.e. the East African Orogen) represent the only potential source of these Pan-African and Grenvillean detrital zircon signatures? Both these zircon age ranges are associated with supercontinent assembly. The Grenvillean and Pan-African zircon sources are widespread in the main continents as shown by Rino et al. (2008) and Voice et al. (2011). Although Antarctica lacks a significant peak in granitoid activity in the 1500 1000 Ma range (Condie et al., 2009), it has been argued from the prominence of the Grenvillean detrital zircons in Early Palaeozoic sedimentary rocks of the Ross Orogen (Goodge et al., 2002, 2004a) and also on the other side of the continent (Prince Charles Mountains) in Permian siltstone that the interior of East Antarctica must include a widespread and significant Grenville source(s) (Veevers and Saeed, 2008, 2011). Grenvillean sources are

16 common in Africa and South America (Rino et al., 2008) and also running across Australia (Myers et al., 1996). The same logic also applies to the Pan-African detrital source which must have been even more widely developed throughout much of East Antarctica (Veevers and Saeed, 2008, 2011). 4.4 Derivation of the Ordovician turbidites from East Antarctica Given that the East African Orogen cannot be considered a unique source for the 1300 1000 Ma and 600 500 Ma detrital signatures found in the Ordovician turbidites, we examine our data in comparison to data presented for other early Palaeozoic sandstones in southeastern Australia. In sample HQ9, two significant clusters occur within the age range 600 480 Ma, one with six grains with ages in the range 510 490 Ma (four of these grains have been checked by duplicate analyses). The other cluster is in the range 560 520 Ma with eight ages. In sample MVP, the youngest significant cluster at 510 480 Ma is based on only two grains (one has a duplicate analysis and the other triplicate analyses). Sample MVP also has a significant cluster in the range 560 520 Ma with 10 ages. Thus both samples show, even if only weakly in MVP, input from a Delamerian detrital zircon signature matching most likely igneous sources in the Delamerian Fold Belt. Cayley (2011) has used the post-collisional volcanic suites of western, central Victoria and Tasmania as a marker horizon (Fig. 9) and these rocks also provide a potential source of the youngest zircons in the Ordovician turbidites. Their relatively restricted distribution in the Ordovician turbidites shows that these volcanic successions were largely buried. Particularly pertinent, is the detrital ages of zircons in the proximal Bilpa Conglomerate of western New South Wales (Fig. 8a) that show most zircons derived within the Delamerian Fold Belt consistent with abundant locally derived metamorphic clasts (see above). A significant population of ~510 485 Ma zircons was noted by Squire et al. (2006) as reflecting a late Cambrian to earliest Ordovician zircon-bearing source in the Adaminaby Group which we would attribute to a Delamerian Fold Belt source. The issue to be resolved is the source of the 600 520 Ma zircons, in particular the 560 520 Ma zircons found in HQ9 and MVP, as well as being widely distributed in other Ordovician and Late Cambrian sandstones (Fig. 8). A likely source is considered to be the now covered continental margin arc of the Ross Orogen of East Antarctica. The northernmost part of this occurs east of the Mertz Glacier in Terre Adelie Land (Fig. 1) and has been shown to include granitic rocks of 580 575 Ma and 515 500 Ma (Goodge and Fanning, 2011). Further south in the Central Trans-Antarctic Mountains, clastic rocks of the Douglas Conglomerate and Starshot Formation include abundant detrital zircons in the range 600 510 Ma as well as a significant Grenville component (Goodge et al., 2002, 2004a). It is clear that the Douglas Conglomerate represents a synorogenic conglomerate derived from the adjacent Ross Mountains and is analogous to the Bilpa Conglomerate of northwestern New South Wales. Additionally, Gibson et al. (2011) noted that another potential source for the 600 500 Ma zircons in the Kanmantoo Group was from erosion of the